Senior ML Performance Engineer

Engineering | Mountain View, United States or Remote | Remote, Remote | San Francisco, United States | Seattle, United States



Working at Atlassian

Atlassians can choose where they work whether in an office, from home, or a combination of the two. That way, Atlassians have more control over supporting their family, personal goals, and other priorities. We can hire people in any country where we have a legal entity. Interviews and onboarding are conducted virtually, a part of being a distributed-first company.

Be the backbone of Atlassian's Agentic AI Integration Products : The Agentic AI Integrations team is responsible for the industry-leading Rovo MCP Server, Agent to Agent integrations as well as on the mission to catapult Atlassian value by leveraging cutting-edge AI capabilities like Claude Skills, ChatGPT/Claude Apps etc., essentially we will be working on anything and everything with AI integrations into the Atlassian ecosystem.

Knack to work on bleeding-edge AI technologies: Passionate to explore and learn AI transformative technologies and quickly pivot from prototyping new initiatives to building highly-scalable enterprise-grade AI products that will be used by 1000s of developers and enterprise users.

ML performance, quality, and systems acumen-ship: Experience in tuning MCP or agent-facing servers for latency, reliability, token efficiency, and tool-selection quality; including dynamic tool discovery, context and response optimization, observability, automated evals, and semantic retrieval using embeddings, vector search, hybrid ranking, and reranking.

• Design, build, and evolve MCP servers, tools, and agent-facing APIs with concise schemas, predictable errors, safe mutations, and clear outcome-oriented contracts.
• Develop accessible, responsive, and performant React and TypeScript experiences that make agent capabilities, MCP tools, and A2A interactions easy to discover, configure, and use.
• Build reusable components, design-system patterns, and frontend architecture that support consistent, scalable user experiences across AI-powered products.
• Integrate GraphQL and REST APIs, SDKs, streaming responses, and real-time data into reliable, user-friendly AI workflows.
• Optimize token and context efficiency through dynamic tool discovery, lazy loading, bounded responses, pagination, selective field retrieval, caching, and reduced tool-call loops.
• Improve end-to-end performance and reliability across front-end clients, gateways, MCP servers, search services, and downstream product systems through observability, tracing, SLOs, and production diagnostics.
• Build semantic retrieval capabilities using embeddings, chunking, vector indexes, hybrid search, metadata and permission filters, ranking, reranking, and freshness strategies.
• Define and operate AI/ML quality programs with JTBD-based evaluations, benchmark datasets, groundedness and relevance metrics, hallucination and bias detection, safety testing, and human feedback.
• Integrate automated evaluations into CI/CD and release gates to detect regressions across model, prompt, tool, and retrieval changes.
• Implement enterprise security and partner cross-functionally to deliver maintainable, well-tested AI integrations, including OAuth 2.1, tenant isolation, audit logging, prompt-injection defenses, and confirmation flows for high-impact actions.